We propose a research program that focuses on the interactions among tick vectors, bacterial pathogens, the community of vertebrate hosts, and the ecological landscapes in which vectors, pathogens, hosts, and humans interact to affect exposure to Lyme disease. We focus on Lyme disease because of its high incidence, widespread distribution, large body of ecological and epidemiological background information, and potential as a model of other vector-borne zoonoses. Our research over the past 9 years has generated a conceptual model that we call the dilution effect. According to this model, high species diversity in the community of hosts for ticks reduces the infection prevalence of ticks by diluting the effects of the ubiquitous white-footed mouse, which is the principal natural reservoir for the disease agent, Borrelia burgdorferi. Other published research suggests that species diversity of vertebrate's decreases, and population density of mice increases, with decreasing forest patch size. We propose to test this model rigorously at sites in six northeastern states. Our research will focus on the generality of the dilution effect and on the mechanisms that underlie it. We will assess the species richness and relative abundance of vertebrate hosts, as a function of size of forest fragments, in suburban landscapes of the Northeast. We will use both species diversity indices and patch size as independent variables in regressions that use either infection prevalence of nymphal ticks, or the abundance of infected nymphs, as dependent variables. Finally, we will use modeling approaches developed in our lab to both assess the generality of the dilution effect for other vector-borne diseases and increase the realism of the Lyme-disease system. Our research comprises a rigorous, multidisciplinary approach to understanding the linkages between ecology and risk of infectious disease.